IMPROVEMENTS IN CLOCK-ESCAPEMENTS. 423 



ment, in the manner which I will explain presently. But first let us consider what are 

 the conditions which such an escapement ought to satisfy*. 



The first of them is that it must be safe from tripping, or from any risk of the 

 pallets by which the scape-wheel teeth are stopped, being thrown so far in lifting that 

 the tooth misses the stop altogether; whereby at least two beats of the pendulum are 

 missed, and perhaps more, and then the scape-wheel acquires such a velocity that several 

 of the teeth may be bent by striking against the stops, and then of course it begins to 

 trip worse than ever. A great many contrivances have been resorted to for preventing 

 this primary difficulty of gravity escapements. Cumming, the first inventor of them, 

 used two pairs of arms, one pair being only for locking and not lifted by the scape-wheel 

 at all, but only by the pendulum in unlocking it. And this was to a certain extent 

 effectual ; and I have heard of one of those escapements going (while it did go) better than 

 a very expensive dead escapement clock, by which it was superseded after it had taken 

 to tripping, from the points of the teeth having gradually got bent by their continual 

 impact against the stops without anything to moderate the velocity. Hardy's escapement 

 was just the same, except that he put all the four arms upon springs, I suppose for 

 the sake of avoiding the friction of so many pivots. But he thereby introduced a worse 

 evil ; for the springs being stiffer in cold than hot weather, and acting on the pendulum 

 at the extremity of its arc, where it is most affected by any variation of force, made 

 the clock gain in winter ; and accordingly Hardy's escapement has gone the way of all 

 other gravity escapements, and was even taken out of the transit clock at Greenwich, and 

 replaced by a dead one. Captain Kater proposed to get rid of tripping by making the 

 impulse arms drop on to an anchor, like the pallets of a dead escapement with the im- 

 pulse faces cut off, and so unlock the scape-wheel by their weight alternately acting on 

 the different sides of the anchor. Mr Gowland, who had a patent escapement in the 

 Exhibition on the same principle as regards the unlocking, provided against tripping by 

 putting paddles to the pallets, going down into a pot of oil which resisted their motion ; 

 not a very elegant contrivance certainly, though apparently effectual. There was also a 

 French escapement by M. Gannery in the Exhibition, very like Gowland's, except that 

 it dispensed with the pot of oil, by giving a large motion of the scape-wheel for a small 

 motion of the arms, as Mr Bloxam's escapement did. But for various reasons not worth 

 enlarging upon now, none of these things have ever come into use, or ever will ; and I 

 know of no other gravity escapements which are not open to still more serious objections. 



But, secondly, it is necessary that the escapement should not only be secure against 

 tripping, but against what may be called, approximate tripping. I have seen escapements 

 in which the depth of the locking was sufficient to keep them from actual tripping with 

 any probable amount of force on the scape-wheel, but nevertheless the arc of the pen- 

 dulum very sensibly increased when the clock-weight was increased. The reason is that 

 the scape-wheel has then force enough to send the arms a little too far, though not far 

 enough to carry the stop quite out of the way of the tooth, and then the pressure of 



* For a description of the general principle of gravity escapemenw I must refer again to my book on Clocks, p. 70. 



